• Post category:Articles
  • Post published:10.7.2024

Pyykkijahti provides knowledge on motivation of crowdsourcing and positioning accuracy of smartphones

How well do gamification and crowdsourcing work for collecting topographic data? In Finland, the potential of crowdsourcing to enhance the cadastral index map of the National Land Survey of Finland (NLS) was investigated with a nationwide smartphone game, Pyykkijahti (Marker Quest). The experiment showed that in some cases smartphone positioning is sufficient to improve the map. Players are motivated to participate in the game not only by getting out into nature, but also by the pleasure of contributing and doing something good.


You are currently viewing Pyykkijahti provides knowledge on motivation of crowdsourcing and positioning accuracy of smartphones
Pyykkijahti (Marker Quest) was a web browser game for smartphones to search for border markers in terrain.

National Land Survey of Finland has studied the potential of crowdsourcing in the digitalized world in many recent projects. The current mundane computing tools from personal computers to smartphones can allow citizens for collecting information up to quality levels of public authorities when collection applications are appropriately designed.

However, crowdsourcing is not a novel method or phenomenon even though digital communities have taken it to a new level: in Finland, for example, long traditions of community volunteering exist in “talkoo”, and globally most influential development projects have always required massive numbers of individuals to strive towards shared goals. In the latest crowdsourcing application of the NLS, the Marker Quest, the capabilities of personal smartphones were studied in the task of enhancing the Finnish cadastral index map.

Smartphone positioning is ready to enhance the map

Interest of the NLS in crowdsourcing originates from the legislative responsibility of maintaining the cadastral index and its founding geoinformation. This is a vast task in Finland that contains almost 17 million border markers indicating the boundaries of the properties.

The NLS is particularly interested in markers positioned less accurately than one metre that are approximated adding up to two-three millions. Border markers have been mounted to terrain since the first land parceling in the 18th century and their thorough traversal by the state surveyors would take decades. Inaccurate locations of the markers is a problem because property boundaries become drawn in wrong positions in the information systems and on maps. Using these wrong positions in terrain potentially causes conflicts at the boundaries. Most inconveniently, a property owner may end up to modifying the terrain on the wrong side of the boundary, for example, by logging.

Marker Quest was a smartphone game designed by the NLS and implemented together with Zoneatlas Ltd. In the game, the user searched for border markers in the terrain and measured their locations using the satellite positioning of the smartphone. The user could also indicate the marker missing, which is valuable information for the NLS as well. Marker quest ran from June 2021 to October 2022 when 4 500 citizens achieved over 30 000 measurements altogether.

Five map images of Finland showing how the number of measurements increased from June to October. Measurements were particularly high in southern Finland and increased throughout the pilot period.
Accumulation of the Marker Quest measurements during the pilot period in 2021.

The NLS evaluated the positioning accuracy of the smartphone measurements in an experiment where smartphone positionings were analysed on 118 professionally measured border markers. The mean positioning error on these markers was 4,18 metres and standard deviation 3,63 metres. However, marker-wise accuracy was found to enhance along with increasing number of averaged smartphone measurements. The results showed that more than 90 % of the markers with over 5 metres positioning error were notably enhanced in location accuracy.

In order to assess the location accuracy of the Marker Quest still further, an accuracy experiment was conducted on precisely measured markers. 41 accurately positioned border markers were measured with about 1 900 positionings using 12 common smartphone models. Accuracy of these measurements was enhanced by repetition and post-calculations based on FinnRef positioning reference network. Finally, the mean of the positioning accuracy reached 1,46 metres and the standard deviation 0,88 metres. Sufficient number of repeated measurements was found to be about 10.

Mean accuracy; >30 measurements per markerUncorrected positioningPost-calculation
Individual measurements9,66 m3,92 m
Marker-wise centre points5,66 m1,46 m
Accuracy of the smartphone positioning with 41 border markers in the accuracy experiment 2022.

Crowdsourcing research reveals community spirit

Successful crowdsourcing requires enthusiastic citizens and particularly so if the conducted task increases in complexity. Marker Quest took the approach of gamification in quite a straightforward manner: playing must be rousing or even fun. With gamification, dull tasks can be turned interesting and entertaining. Gamification methods exist in many forms, most familiar and straightforward being probably collecting points and keeping a scoreboard. These two were used in Marker Quest for uplifting the marker search.

When Marker Quest players were asked to describe themselves in a questionnaire, about 400 respondents felt to be mainly “altruists” and “builders”. The players felt themselves being least “socializers” and “profit chasers”. That is, doing good was significant for the players and probably made them search one marker more.

Distribution of player types in the Marker Quest.

Based on the results from the Marker Quest, future research topics contain reasons for variating positioning accuracy of smartphones as well as accuracy differences between smartphone models. As for the crowdsourcing, more insights are needed about how players can be motivated to repeat measurements and make them longer for still enhancing the accuracy.

Matko and Luore development and research projects of the NLS were lead by Heikki Lind (steering groups) and Juha-Matti Vääränen (project groups).


Research article: Rönneberg, M. and P. Kettunen, 2023. A gamified map application utilising crowdsourcing engaged citizens to refine the quality and accuracy of cadastral index map border markers. International Journal of Digital Earth, 16(2), 4726-4748. https://doi.org/10.1080/17538947.2023.2279673

Master’s thesis: Jussila, A., 2023. Positioning accuracy of smartphones in crowdsourcing context. Master’s thesis, Aalto University. http://urn.fi/URN:NBN:fi:aalto-202305213319

Bachelor’s thesis: Kontiokoski, A., 2022. Enhancing Location Accuracy of Boundary Markers by Crowdsourced Smartphone Positioning (in Finnish). Bachelor’s Thesis, Land Surveying, Lapland University of Applied Science. https://urn.fi/URN:NBN:fi:amk-202202252860